Quantifying the calibration uncertainty attributable to thermocouple inhomogeneity

Inhomogeneity in the Seebeck coefficient as a function of position along a thermocouple wire frequently dominates the uncertainty budgets of thermocouple calibration and use. The calibration process itself, simply through exposure to elevated temperatures for relatively modest times, generates both reversible and irreversible changes to the thermocouple that are a complex function of time, temperature, alloy composition, sheath structure, etc. We present data acquired using a salt bath at 250 °C to provide the step-function-like gradient that is our spatial probe of thermoelectric homogeneity. We show how the finite width of the step-function limits our ability to assess the "true" inhomogeneity of the thermocouple, and explore how the inhomogeneity impacts the calibration uncertainty attainable with the various thermal sources used for the calibration of thermocouples (based on their characteristic temperature gradients).